Synthetic and Network Assisted Digital Twin Platform for Signal Emulation (SYNAPSE)

AiRANACULUS, in association with the University of Texas, proposes to develop Synthetic and Network Assisted Digital Twin Platform for Signal Emulation (SYNAPSE), which will meet and exceed the requirements of the “Artificial Intelligence / Machine Learning (AI/ML) Ready Synthetic Radio Frequency Data” project. Technical Objective for the SYNAPSE Project is to develop improved techniques for generating and labeling synthetic data across different classes of RF signals. This data will be used to train Electronic Support (ES) and Signals Intelligence (SIGINT) models, enhancing the automation of detecting, characterizing, and identifying Signals of Interest (SoI). Through Artificial Intelligence (AI) and Machine Learning (ML), this effort aims to tackle the growing volume and complexity of RF signals that traditional methods cannot fully address. The innovation seeks to lessen operator workload and boost battlespace awareness and decision-making capabilities in both strategic and tactical Army operations.                   The rapid growth in RF signals has created a substantial challenge in maintaining situational awareness of unit and system surroundings.  Traditional methods for identifying SoI linked to potential threats, targets, or even tracking blue force and civilian activities are struggling to keep up. This surge in RF technology has been fueled by the widespread availability and cost-effectiveness of Software-Defined Radios (SDR) and new communication protocols supporting the Internet of Things (IoT) and related networking infrastructure. To manage this expanding signal landscape, automated detection and characterization are essential. AI / ML are central to this automation, along with large volumes of AI-ready data to train and refine the models that will perform these tasks. However, measured data collections are often costly, time-intensive due to the complexities of coordinating in dense battle environments and carry risks due to the number of involved elements. As a result, synthetic data has become a critical component of the Army's data strategy.                   We anticipate the SYNAPSE Digital Twin to be a standalone self-sufficient Dockerized Software Module that can be delivered to the Army for integration with the core components of Project Linchpin. The SYNAPSE Systems Architecture has following core components: 1. Transmit and Interference Waveform Library. We plan to leverage this Module from the IARPA SCISRS Program. 2. Transmit Waveform and Interference Synthesis, 3. Active Learning, 4. Generative AI Module, 5. Nvidia Sionna Emulation Engine, 6. Data Augmentation, 7. Addition of Hardware Artifacts (OTA Component), 8. Synthetic Data Verification and Validation using Performer Algorithms, and 9. Synthetic Data Scoring. In the following sections, we provide detailed description of all these core components of SYNAPSE.